VituixCAD v2

[Reet]

Users have been asking do they need measurement data in vertical plane, negative angles, smaller than 10 deg steps etc. ASR Directiva project has measurement data for some conclusions, applicable to circular radiators in horizontally symmetrical small box.
Solid lines = both planes and negative angles, step 5 deg, totally 143 responses / driver.
Dash-dot lines = positive angles in horizontal plane only, step 10 deg, totally 19 responses / driver.

I thought the need or lack thereof with vertical angles and negative angles was already fairly well established. There are use cases for vertical data, such as ribbon or AMT drivers. I detailed a simplified method of approximating the vertical data for a driver without measuring in the following thread, unfortunately there didn't appear to be any interest in this.

http://www.htguide.com/forum/showthread.php?44717-Simplified-polar-data-generating-in-VituixCAD

[Reet]

Same merge comparison as above, this time using the "cardioid" text book function. This time, the text book function is proving decreased power response when compared to my measured data, so the measured data of a 120mm driver in a cabinet is falling somewhere in between cardioid and omni, which if it's true, maybe there needs to be another selection in the drop down...quasi-cardioid or something like that in between

[Efalegalo]

I thought the need or lack thereof with vertical angles and negative angles was already fairly well established. There are use cases for vertical data, such as ribbon or AMT drivers. I detailed a simplified method of approximating the vertical data for a driver without measuring in the following thread, unfortunately there didn't appear to be any interest in this.
http://www.htguide.com/forum/showthr...g-in-VituixCAD

Hi Reet,

I should have posted there. I found the write-up very helpful. Thank you.

[kimmosto]

I thought the need or lack thereof with vertical angles and negative angles was already fairly well established.

Yes, but (mostly) in theory. That's maybe the first public and quite reliable data set as actual measurements in both planes with 5 deg angle step. Anyway, result is a bit surprising, but within predicted/acceptable tolerance so vertical measurement data and 5 deg angle step can be useful and informative but not mandatory with circular hifi drivers.

I detailed a simplified method of approximating the vertical data for a driver without measuring in the following thread, unfortunately there didn't appear to be any interest in this.

Low interest is common nowadays. I don't usually simulate and recommend simulation other than baffle step with Diffraction tool.

[Reet]

Thanks for presenting the comparative data. It’s good to have some context to the potential error in simulation that results from time saving measures.

In my case of the AMT tweeter, the error that resulted from generating off-axis data using the diffraction tool versus using real measurements was rather small, the diffraction tool may be more accurate than you think . Of course information like breakup pattern of a woofer won’t be captured, but the piston model used gets us most of the way there. I don’t have a turntable for off-axis measurements, just a laminated polar grid that I tape to the floor, the process is rather tedious and I decided that providing the vertical data by diffraction sim was “good enough” to save me the work of taking those measurements. The intent is to provide the options, and with an understanding of the error in the result the user can decide if a simplified process is good enough, or to take the extra measurements in pursuit of perfection.

[Reet]

Playing around with my "dipole" speaker today, I've decided that viewing the polar map as a full 360 degree display is a bit awkward. I think for the purpose of viewing dipole speaker polar data, it would be great to have an additional toggle when the "show +/- 90 deg" option is selected, in order to switch the view from the front hemisphere to the rear hemisphere.

[kimmosto]

ANSI/CTA-2034-A Final:

5.2 Post Processing of Data
The following composite response curves shall be calculated. In each instance a power average of the specified magnitude responses shall be calculated.

Listening Window
The listening window curve is a spatial average of the nine magnitude responses in the ±10º vertical and ±30º horizontal angular range.
• 0°
• ± 10º vertical
• ± 10º, ± 20º, ± 30º horizontal

Early Reflections
The early reflections curve is an estimate of all single-bounce, first-reflections, in a typical listening room.
• Floor Bounce: 20º, 30º, 40º down
• Ceiling Bounce: 40º, 50º, 60º up
• Front Wall Bounce: 0º, ± 10º, ± 20º, ± 30º horizontal
• Side Wall Bounces: ± 40º, ± 50º, ± 60º, ± 70º, ± 80º horizontal
• Rear Wall Bounces: 180º, ± 90º horizontal

Vertical Reflections
The “floor reflection” is defined as the spatial average of three measurements at 30 degrees below the main-axis ± 10°. The “ceiling reflection” is defined as the spatial average of three measurements at 50° above the main-axis ± 10°.
• Floor Reflection: - 20°, - 30°, - 40° vertical
• Ceiling Reflection: + 40°, + 50°, + 60° vertical Horizontal Reflections
• Front: 0°, ± 10º, ± 20º, ± 30º horizontal
• Side: ± 40°, ± 50°, ± 60°, ± 70°, ± 80° horizontal
• Rear: ± 90°, ± 100°, ± 110°, ± 120°, ± 130°, ± 140°, ± 150°, ± 160°, ± 170°, 180° horizontal, (i.e.: the horizontal part of the rear hemisphere).

How to interpret the text above.
1) What responses are "power averages" calculated as R.M.S of sound pressures?
2) What responses are "spatial averages" calculated as arithmetic mean of sound pressures?

Background is that there's small difference between VituixCAD and Klippel - especially in ER and ERDI. I suspect that VituixCAD uses too much RMS instead of mean, but standard is far too ambiguous at least for me. It would be very nice to get 100% certain answer from member of standard committee.

[Reet]

ANSI/CTA-2034-A Final:


How to interpret the text above.
1) What responses are "power averages" calculated as R.M.S of sound pressures?
2) What responses are "spatial averages" calculated as arithmetic mean of sound pressures?

Background is that there's small difference between VituixCAD and Klippel - especially in ER and ERDI. I suspect that VituixCAD uses too much RMS instead of mean, but standard is far too ambiguous at least for me. It would be very nice to get 100% certain answer from member of standard committee.

I'm not sure who you would even reach out to for clarification on the standard, it only references the "R3 Audio Systems Committee", guess you have to become a member to know more.

https://standards.cta.tech/kwspub/home/Committees/

It looks like for $500 annually you can join the CTA standards committee as a non-member. Unfortunately it looks like in order to become a CTA member you must have a business with a location in Canada or USA.It also looks like there was a CTA-2034-B standard in development, but it's been put on hold as of a year ago.

https://standards.cta.tech/apps/group_public/project/details.php?project_id=540

FWIW I don't lean too heavily on ER or ERDI, but rather the in-room response and power response is of greater importance to me.

[kimmosto]

PIR = 0.12 * LW + 0.44 * ER + 0.44 * SP so average types of both listening window and early reflections should be okay to get correct PIR result. Difference between arithmetic mean and RMS is quite small so possible error in ER and ERDI should be max. 0.5 dB. Problem is that both CTA-2034-A and 'Sound reproduction' by Toole don't specify calculation clearly enough. Average types in VCAD are based on second hand information on diyaudio.com. Klippel may or may not know better, but VCAD is more suspicious as long as programs give different result.

Also power response (SP) calculation has small difference compared to CTA-2034-A. It's usually visible as a small difference in tilt of power and DI responses. Weighting tables in ANSI/CTA-2034-A seem to have some issues. Table for 10 deg angle step does not give correct DI with ideal gradient radiators, and table for 5 deg angle step is totally screwed up. So VituixCAD continues radial->spherical conversion with close to integral sine function (or Archimedes' Hat Box Theorem) which is theoretical ideal with angle step close to 0 deg. It's also easy way to support other than constant 5 or 10 deg and variable angle steps and angle coverage.
VCAD would be able use weighting values in 10 deg table with any constant angle step, but that feature is not in use at the moment to avoid change in the result if user switches to variable angle step (which is not supported by CTA-2034-A).

Anyway, I have already changed averaging of listening window, floor, ceiling, front, side and rear early reflections from RMS to arithmetic mean. Also ER ver, ER hor and ER tot can be changed to mean, but I don't dare to publish anything before correct interpretation is available.

[fluid]

How to interpret the text above.
1) What responses are "power averages" calculated as R.M.S of sound pressures?
2) What responses are "spatial averages" calculated as arithmetic mean of sound pressures?

Hard to know exactly what was intended given the lack of clarity in the terminology, but the first line seems quite clear to me

In each instance a power average of the specified magnitude responses shall be calculated,
I don't understand how that could be interpreted as anything other than saying each named curve is a power average of the indicated individual curves which suggests RMS.

If it was legislation that would be my argument anyway

[kimmosto]

I don't understand how that could be interpreted as anything other than saying each named curve is a power average of the indicated individual curves which suggests RMS.

Yes, but why 'spatial average' is mentioned right after that at least for listening window and both vertical reflections? If we believe www, spatial average is arithmetic mean value over multiple points in space. Not RMS. Also Toole's book says nothing about power or RMS. Just spatial average or plain average until Sound power which is "total acoustical energy radiating through an imaginary spherical surface ...". In addition, simple average can be calculated from pressure or dB values. Using dB values is quite common though not much sense at least for averaging.

My opinion is that RMS of pressure values is the best for this kind of benchmarks no matter what gurus and standards try to say. This would lead to statement that VituixCAD is the reference

[fluid]

Yes, but why 'spatial average' is mentioned right after that at least for listening window and both vertical reflections? If we believe www, spatial average is arithmetic mean value over multiple points in space. Not RMS.

That is the confusing part because the first line is much clearer, it could just be that they are using the word spatial to indicate multiple separate positions and nothing more.

This old acoustics paper suggests that a squared pressure average is more accurate for sound power in reverberant spaces.

To measure sound power in a reverberation room, we need to estimate the space‐time average of squared sound pressure. While the time average usually presents no special problem, the space average may, particularly for acoustic signals having pure‐tone or extremely narrow‐band spectra. The space average is subject to a random error due to finite sample size. Tables and formulas are presented from which an experimenter may determine the confidence level of his estimate in terms of signal‐bandwidth reverberation time. and number of fixed microphones or the path length for a moving microphone. The space average is also subject to a bias error if the quantity averaged is sound level or rms pressure instead of squared pressure.

https://asa.scitation.org/doi/10.1121/1.1972148

Maybe that might be worth a try?

My opinion is that RMS of pressure values is the best for this kind of benchmarks no matter what gurus and standards try to say. This would lead to statement that VituixCAD is the reference

That could be true

[kimmosto]

Playing around with my "dipole" speaker today, I've decided that viewing the polar map as a full 360 degree display is a bit awkward. I think for the purpose of viewing dipole speaker polar data, it would be great to have an additional toggle when the "show +/- 90 deg" option is selected, in order to switch the view from the front hemisphere to the rear hemisphere.

Rotating whole construction programmatically would be very easy solution, but result is illogical because everything including on-axis and listening window is rotated 180 deg in hor plane. Splitting and mirroring angle axis of polar map, waterfall and surface chart may be doable, but complicate. Re-ordering of chart series is possible, but also that would require some tricks to angle axis of waterfall and polar map.
Rear hemisphere of dipole is clearly and logically visible as polar map, waterfall and surface chart, and angle values stay logical. Just show full space and tilt waterfall and surface if needed and that's it. Paint is available if 180 deg must be in the middle of polar map

[kimmosto]

Maybe that might be worth a try?

That's for sound power in a reverberation room. Spinorama is calculated from anechoic responses which very simple. We just need inside information or new revision of the standard with clear definitions.

[fluid]

We just need inside information or new revision of the standard with clear definitions.

Given that it has only recently been revised the second seems unlikely.

DonVK made a VACS calculation example of 2034 and he chose rms average too. I changed the calculation to arithmetic mean, the difference seems pretty small.
Blue is RMS red is mean, Listening Window.



I wonder if I simulate a perfect point source and see if there is a difference in SPDI between the different methods of averaging the listening window? The sound power is a weighted rms average with the weightings given so that at least seems clear. SPDI should be 0 according to the standard with a perfectly omnidirectional source.

[kimmosto]

Of course, drivers in a cabinet aren't completely omni even at 20Hz, so maybe the cardioid text book function would be a more fitting comparison here.

Change in sound travel distance from 0 deg to 180 deg causes weak directivity to finite distance. Vent at rear panel causes weak dipolic feature slightly below and above tuning frequency. Diffraction increases directivity to 100...170 deg, but also this is small at 20 Hz with typical boxed speakers. Very far from cardioidish with DI=4.8 dB.
Anyway, with the latest revision you can add some small directivity also to omni. For example monopole portion of 94 % produces DI=0.5 dB.

Rev. 2.0.76.3 (2021-09-04)

Merger
* 'Polar pattern' list box replaced with 'Force to Gradient' checkbox and 'Monopole portion' text box.

[Reet]

Thank you, yes approximation if real world monopole behaviour would be my goal for this feature so this change is very welcome.

Have you given any thought to my suggestion of being able to view both front 180 degree and rear 180 degree polar maps when the “show +/- 90 deg” option is selected? I think there is benefit to this display when observing dipole systems.

[kimmosto]

Have you given any thought to my suggestion of being able to view both front 180 degree and rear 180 degree polar maps when the “show +/- 90 deg” option is selected?

I struggled few days with it, but not success. Perhaps the brain suffers from a lack of oxygen. Answer is here.

[Reet]

I struggled few days with it, but not success. Perhaps the brain suffers from a lack of oxygen. Answer is here.

I'm definitely the one lacking oxygen today, as I skipped right over your response above. It sounds like it is a complicated task to program, so don't worry about it. I have a "band-aid" solution to view the plot, which is to simply flip the angles in the driver configuration, then take a screenshot to compare to the front orientation. It's interesting to see how much the driver geometry affects the response pattern from front and back sides of a dipole.

I have only one more request with the polar maps, hopefully this is a simple solution. Generally, when viewing the polar map the colour red is represented as the same SPL as the top of the vertical axis on the SPL and power charts. However, when I switch between crossover variants, I see that the colour SPL range on the polar map gets adjusted, it must be doing some automatic scaling based on the system response. I then have to change the SPL scale in order to get the polar map colours to line up again. It would be great to not have to do this, I would love to be able to switch back and forth between crossover variants and see the difference in the polar map immediately, but this is not easily done if the scaling changes when I do so. Basically, if the top of the vertical scale for the SPL chart is 90dB, then the colour red for the polar map should be 90dB always.

[kimmosto]

Rev. 2.0.76.4 (2021-09-05)

Main
* Added 'Rotate 180 deg' checkbox to Room tab.
* Scaling of unnormalized polar map and surface chart fixed when XO variation is changed.

[Reet]

Rev. 2.0.76.4 (2021-09-05)

Main
* Added 'Rotate 180 deg' checkbox to Room tab.
* Scaling of unnormalized polar map and surface chart fixed when XO variation is changed.

You're the best

[Evil Twin]

You're the best

Agreed!

[tktran]

Me three!

FAQ #253

Btw, are you planning on adding Dipole modelling to the Enclosure tool?

[kimmosto]

^Dipole model = Infinite baffle with Enclosure + Open baffle with Diffraction tool.

[tktran]

h:

Does the shape of the frame matter? Eg. H, U, W, N frame

[kimmosto]

Does the shape of the frame matter? Eg. H, U, W, N frame

Diffraction tool simulates I-frame. Other types are dipolar transmission lines so you need something else. Symmetrical designs are close to I-frame at very low frequencies. Response depends on distance between rear and front wave.

[Quza]

Smallish bug report: The red X in the top right is not entirely in the top right, so when trying to close VituixCAD by flicking the mouse to the top right corner and clicking, it often closes programs that aren't even in focus.
(Win10x64)

[kimmosto]

^You should report Windows, .NET and display driver issues directly to Microsoft or manufacturer of virtual machine player in case your session is not running on real machine. I have nothing to do with those.

[kimmosto]

2.0.76.6 (2021-09-26)

Main
* Added Normalized command to context menu of GD & Phase chart. Selected (clicked to bold) phase response is normalized to 0 deg line.

[Reet]

Simple request, allow the angle field in the calculator to be an editable field for manual entry. Sometimes the calculator incorrectly grabs a driver model number from the file name as the angle and the only way to fix is to rename my files. I'd also prefer that the scale and delay values are rest to zero when the calculator window or program is closed.

[kimmosto]

^Done in the latest build a minute ago. Revision number not changed.

[Reet]

Thanks!

[Reet]

Something seems a bit off with the vertical scale of the resonant frequency chart on the box volume calculator window.

What I get:


What the help file shows:

[kimmosto]

2.0.76.7 (2021-10-11)

Auxiliary calculator
* Y scale, Q factors and range of resonance chart improved.

[Reet]

Better

[kimmosto]

I don't remember why curves had gain=2, but it did not look a mistake. Maybe curves were cropped because peaks with Q=10 and linear Y scale don't look smooth with 1/48 oct frequency resolution. Another solution would be thick vertical lines.

[Reet]

Minor details, I just thought it looked odd to crop off the top of the chart. I’m good either way, I think I have an old spreadsheet that displayed this type of information with simple vertical lines, but the high Q peaks are probably more representative of reality.

[kimmosto]

Ancient tiny bug fixed. This was not noticed because I never use automatic snap to E12/E24/E48 series with Optimizer.

2.0.76.8 (2021-10-16)

Main, Optimizer
* XO network recalculated after rounding with E12, E24 or E48 snap.

[Efalegalo]

A quick question about using VituixCAD with a slanted baffle (I'll have the moderator delete if this should really be in a separate thread):

I have built a speaker that tilts back at 11 degrees. When I took polar measurements in Arta, I took them with a wedge underneath the speaker that restores the speaker back to 0 degrees. I took measurements of each driver (there are two: tweeter and woofer) with the elevation of the mic at the center point of the driver under test.

With the measurements performed as I described above, how I do now correctly specific diver placement in VituixCAD?

My "Design Axis" is tweeter level. As indicated above, the speaker is tilted back 11 degrees. Looking at a straight baffle (i.e., no-tilt), the woofer is placed exactly 220mm below the tweeter, but also in the center of the baffle (horizontally).

Is it simply:

Tweeter (x=0,y=0,z=0,r=0,t=11)
Woofer (x=0,y=-220mm,z=0,r=0,t=11)

[kimmosto]

^Tweeter X=0, Y=0, Z=0, R=0, T=11
Woofer X=0, Y=-216, Z=-42, R=0, T=11

[Efalegalo]

^Tweeter X=0, Y=0, Z=0, R=0, T=11
Woofer X=0, Y=-216, Z=-42, R=0, T=11

Got it! I see how it's being calculated. Thank you.

[augerpro]

Question for Kimmosto. I was in a discussion about ctc distance between a tweeter and mid, and it was said that you had shown 1.0-1.2 wavelength at crossover was better for power response than say, .7 wavelength. Can you explain the work you did to come up with this? What were the assumptions?

[Reet]

Question for Kimmosto. I was in a discussion about ctc distance between a tweeter and mid, and it was said that you had shown 1.0-1.2 wavelength at crossover was better for power response than say, .7 wavelength. Can you explain the work you did to come up with this? What were the assumptions?

Some posts saved from the DIY Audio discussion may be of some help to you, to avoid repetition.

1) Waveguides are good, they make the tweeter less directive at XO than the woofer.

Possibly but not necessarily. Wave guide is quite large - possibly too large in many modern speakers though large wave guide might be needed with shoe box enclosure. Equal directivity at XO causes hump to directivity index without significant (~90 deg) phase mismatch.

2) Usually I try to make the tweeter and woofer as close together as possible. Are you saying there is a specific formula for spacing? And that maybe closer isn't always better?

"c-c distance ca. 1.2 x wave length at XO."
Could you expand on this?
'As close as possible' could be 'the worst possible' for directivity index i.e. either on axis (~listening window) or power response or both should be compromised to get balanced sound.
Of course if minimal vertical lobing is priority #1 then you should locate as close as possible. Coaxial driver wins that game always, but otherwise not necessarily...probably.

With simplified theory c-c = 1/2 wave length is the worst case for power response with equal DIs, and c-c = wave length at XO is the best case. Simply because sum with difference of 1/2 wave length is null and vertical +/-90 deg have the biggest weight in power calculation (due to dual orbit data to spherical intensity conversion). Early vertical reflections have significance too and DI of different radiators are not always equal => the smoothest DI and ERDI is found when c-c = 1.0-1.4 x wave length. This means that possibility of the worst DI is when c-c = 0.5-0.7 x wave length.

c-c studies are ridiculously easy with VituixCAD. Just load measurement data of the radiators, create ideal flat on axis response (with Optimizer and G(f) blocks) with estimated XO and tune driver's Y mm until combination of DI and ERDI is the best.

If I were to elaborate a bit, I think the common rule of thumb of "closer is better" is a misconception that originated from limited information. If you were to observe only the polar pattern generated by the crossover, you will probably come to the conclusion that closer together is better as it creates a wider vertical polar lobe. The polar plot function in SoundEasy is a perfect example of this. However, this wide lobe is not providing accurate information of in-room acoustic performance. Observing the in-room and power response plots in VituixCAD provide better insight of acoustic performance, and at 2-3m listening distance, a bit of difference in vertical polar lobe is no longer important at all.

The following post may be of some help as well, related to directivity control.

3) Just your regular roundover or chamfer, the larger the better. How different are roundovers vs. a chamfer? Can you model chamfers with the diffraction tool?

What about square edges around the woofer and large "facets" cut around the tweeter instead of the larger roundovers/chamfer?
Quite many questions and I don't know what "facet" is.
I prefer rounded chamfers, 45 deg and R>=32mm, both quite easy to manufacture. Tweeter should have very small effective baffle size to create directivity above typical XO frequency.

Diffraction tool is limited, but supports designing flat baffle area so that directivity dips and humps due to edges compensate directivity dips and humps of drivers and estimated combination of them assuming phase match at XO. Few tips:
- Do not increase directivity at XO with the box because phase matched XO does it anyway.
- Do not increase directivity below XO with the box because woofer cone does it anyway.
- Increase directivity above XO with the box if tweeter does not have wave guide or wave guide is small.

[augerpro]

I was hoping for actual measurements or at least sims, the above are all just claims/statements. The closest thing that speaks to the assumptions is the part about "vertical +/-90 deg have the biggest weight in power calculation". I'd like to understand this.

[Reet]

I was hoping for actual measurements or at least sims, the above are all just claims/statements. The closest thing that speaks to the assumptions is the part about "vertical +/-90 deg have the biggest weight in power calculation". I'd like to understand this.

Vertical response varies in a greater manner with crossover frequency and driver separation than most any other factor, so of course it will have significant influence on power response as well. You don't have to take these "claims" at face value, just measure any of your drivers and load the data into VituixCAD and analyze the results for yourself.

"c-c studies are ridiculously easy with VituixCAD. Just load measurement data of the radiators, create ideal flat on axis response (with Optimizer and G(f) blocks) with estimated XO and tune driver's Y mm until combination of DI and ERDI is the best."